-
The Journal of Neuroscience, November 1987, 7(11): 37493783
Phenotypic Properties of Catecholamine-Positive Cells That
Differentiate in Avian Neural Crest Cultures
Douglas S. Christie,” M. Elizabeth Forbes, and Gerald D.
Maxwell
Department of Anatomy, University of Connecticut Health Center,
Farmington, Connecticut 06032
We have investigated several phenotypic features of the
catecholamine-positive (CA+) cell population that develops in quail
neural crest cultures. The number, spatial distribu- tion, and
morphology of CA+ and tyrosine hydroxylase-pos- itive (TH+) cells
are similar at all ages examined, suggesting that these 2 cell
classes are identical. Neither CA+ nor TH+ cell bodies or processes
were stained using antisera that recognize the 70 or 160 kDa
subunits of chicken neurofila- ment protein. Other cell bodies and
fibers in the cultures (which were CA- and TH-) were stained with
these neu- rofilament antisera. The uptake and storage of
3H-norepi- nephrine by neural crest cultures containing CA+ cells
were inhibited in the presence of desmethylimipramine and by
incubation at OX, but were unaffected by normetanephrine. Overnight
treatment with reserpine eliminated histochemi- tally detectable CA
fluorescence from the cultures. Chronic reserpine treatment from
day 2 to 7 in vitro prevented the appearance of CA+ cells, while
normal numbers of TH+ and somatostatin-like immunoreactive (SLI)
cells developed. The number and light-microscopic morphology of the
CA+ cells that developed in these cultures were not dramatically
al- tered by either exogenous NGF or 6-hydroxydopamine. Us- ing the
method of Grill0 et al. (1974), we have demonstrated that the CA+
cells observed in the light microscope corre- sponded to cells
containing abundant cytoplasmic granular vesicles (GV)
characteristic of catecholamine storage gran- ules observed in
other systems. The GV diameters were quite similar in cells
examined after 5, 7, 14, and 21 d in vitro. Most GV were 50-200 nm
in diameter and were dis- tributed in a unimodal manner, with the
observed modal val- ues in the range of 85-115 nm at the ages
examined. The number of GVlrm* of cytoplasmic area remained quite
con- stant at all ages examined. These data, taken together with
other available information, suggest that the CA+ cells that
differentiate in our neural crest cultures resemble, in many
respects, the small, intensely fluorescent cells found in auto-
nomic ganglia and extra-adrenal chromaffin tissue of many species.
At present, we do not know if the CA+ cells that
Received Feb. 1 I, 1987; revised May 7, 1987; accepted May 14,
1987. This work was supported by NIH Grants NS16115, Research
Career Devel-
opment Award NS00696 (G.D.M.), and NRSA NS 17841 (D.S.C.). We
would like to thank Drs. Gudrun Bennett, Hiroshi Hatanaka, and Bill
Tank for their generous gifts of antibodies, Dr. Doug Oliver for
the use of his equipment and helpful discussions, and Dr. Michael
Rosenberg for performing the NGF RIA determinations and commenting
on the manuscript.
Correspondence should be addressed to Dr. Gerald D. Maxwell at
the above address.
a Present address: Anatomy and Physiology Branch, Medicine and
Surgery Di- vision, Academy of Health Sciences, Fort Sam Houston,
TX 78234-6100.
Copyright 0 1987 Society for Neuroscience 0270-6474/87/l
13749-15$02.00/O
differentiate in our neural crest cultures are a stable endpoint
of development or whether they are a developmental inter- mediate
in adrenergic differentiation that is normally ob- served only
transiently during the development of avian sym- pathetic ganglia
in viva, but that can persist under our tissue culture
conditions.
The embryonic neural crest is the source of the progenitor cells
that migrate, proliferate, and differentiate to form most of the
peripheral nervous system, including the neurons of the sym-
pathetic, parasympathetic, enteric, and most sensory ganglia
(Weston, 1970; Noden, 1978; LeDouarin, 1982). In addition, neural
crest cells are the source of many other cell types, in- cluding
the cells of the adrenal medulla, Schwann cells, melan- ocytes, and
skeletal and connective tissue of the head and face. We have been
studying the neuronal differentiation of neural crest cells in
tissue culture, with particular emphasis on those cells that
develop adrenergic properties.
During embryonic development, neural crest cells acquire
particular differentiated traits that can be used as phenotypic
markers of their maturation into neurons. One such marker is the
acquisition of the capacity to metabolize specific neuro-
transmitter compounds. A subset of neural crest cells grown in
tissue culture develops into cells with the capacity to synthesize,
store, take up, and release catecholamines (CA) (Cohen, 1977;
Sieber-Blum and Cohen, 1980; Fauquet et al., 1981; Maxwell et al.,
1982; Maxwell and Sietz, 1983, 1985). Some of these neural crest
cells that become CA+ also develop somatostatin- like
immunoreactivity (SLI) (Maxwell et al., 1984a; Sieber-Blum, 1984).
This differentiation of neural crest cells in tissue culture
parallels in many respects their development in viva In avian
embryos, CA and SLI appear in neural crest cells shortly after they
have ceased their initial migration and coalesced to form the
primary sympathetic trunks (Enemar et al., 1965; Kirby and Gilmore,
1976; Rothman et al., 1978; Garcia-Arraras et al., 1984; Maxwell et
al., 1984b).
While the presence of CA is a useful phenotypic marker, it is
not sufficient for specifying precisely what type ofCA-containing
cell is present (Black, 1982). There are 3 major classes of CA-
containing cells derived from the neural crest that are found in
adult animals. These are principal sympathetic neurons, small,
intensely fluorescent @IF) cells, and the chromaffin cells of the
adrenal medulla. In addition, CA+ cells derived from the neural
crest are found in the carotid body and in paraganglia located in
various sites (Biscoe, 197 1; Mascorro and Yates, 1980). Also, in
some species a small population of CA+ cells may exist in adult
sensory ganglia (Price and Mudge, 1983). During embry- onic
development, transient CA+ cells are present in the gut
-
3750 Christie et al. * Adrenergic Neural Crest
Differentiation
and dorsal root ganglia of rat and mouse embryos (Cochard et
al., 1978, 1979; Teitelman et al., 1978, 1979; Jonakait et al.,
1984). Although principal neurons, SIF cells, and chromaffin cells
all contain CA, they exhibit very significant differences in terms
of their structures, responses to growth factors and hor- mones,
and physiological functions. One major difference is the type and
distribution of CA storage granules. In principal neu- rons, CA
storage granules are about 50 nm in diameter, are found almost
exclusively in cellular processes, and are virtually absent from
the cell body (Gtillo, 1966; Bumstock and Costa, 1975; Gabella,
1976). In SIF cells and chromaffin cells, the CA storage granules
are larger and are abundant in cell bodies (Er- anko, 1975). SIF
cells and chromaffin cells can be further dis- tinguished from each
other by the size of the CA storage vesicles and the presence of
processes in some SIF cells (Coupland, 1965; Grille, 1966; Taxi,
1979; Taxi et al., 1983). A second difference is in the response of
these cell types to growth factors and hor- mones. During embryonic
development, principal sympathetic neurons require NGF for their
survival and differentiation (Levi- Montalcini and Angeletti, 1968;
Greene and Shooter, 1980; Thoenen and Barde, 1980). Adrenal
chromaffin cells and SIF cells show dramatic responses to
glucocorticoid hormones with respect to survival and
differentiation (Doupe and Patterson, 1980). However, in addition,
both SIF cells and adrenal chro: maffin cells from neonatal rats
can be influenced by NGF to acquire principal neuron-like traits
(Unsicker et al., 1978; An- derson and Axel, 1985; Doupe et al.,
1985a, b).
monia vapors (Ehrmann and Gey, 1956). In the experiments
reported here, the neural tubes were retained in the cultures. NGF
was not added to the the growth medium except where specifically
indicated.
Immunocytochemistry and histochemistry. Indirect immunofluores-
cence was used to visualize SLI-positive cells, as described
previously (Maxwell et al., 1984a). Cells that were TH+ were also
visualized by indirect immunofluorescence using the same protocol
as for SLI, except that the fixation time was reduced to 40 min.
Antiserum directed.against TH was the generous gift of Dr. A. W.
Tank and was used at a final dilution of 1: 100.
Histochemistry to reveal CA+ cells was performed using the
method of Fumess et al. (1977) which results in a water-stable
fluorophore. Cultures were washed once with Hanks’ balanced salt
solution (HBSS)- HEPES, followed by incubation in 4%
paraformaldehyde and 0.1% glutaraldehyde for 2 hr at room
temperature. Cultures were then washed in PBS and mounted using
glycerol : PBS (3: 1, vol/vol). For neurofila- ment staining in
conjunction with CA+ cell visualization, the cultures were
processed to reveal CA as described above, followed by indirect
immunofluorescence for neurofilaments, as described by Bennett et
al. (1984). This allowed CA+ cells to be observed
using-catecholamine outics and neurofilament staining using
tetramethvlrhodamine (TRITC) optics. Control experiments
demonstrated that there is no overlap be: tween CA and TRITC
optics. Antisera raised in rabbits directed against the 70 kDa
(NF70) and 160 kDa (NF160) chicken neurofilament pro- teins were
generously provided by Dr. G. Bennett and were used at dilutions of
1: 150 and 1500, respectively. These antisera have been shown to be
specific for the neurofilament protein they were raised against,
and showed no cross-reactivity with each other (Bennett et al.,
1984).
For visualization of TH-positive and neurofilament-positive
struc- tures in the same experiment, double-label indirect
immunofluorescence was used. Monoclonal antibody to TH was the gift
of Dr. H. Hatanaka and was used at a dilution of 1:200. The
neurofilament antisera were those noted above. Cultures were either
fixed in acetone at -20°C for 10 min or in 4% paraformaldehyde for
40 min. Both fixation protocols gave the same results. Cultures
were grown on collagen-coated glass coverslips when acetone was
used. Control experiments using the “in- correct” primary
antibodies were performed to demonstrate that the
fluorochrome-conjugated secondary antibodies possessed the required
specificities. At the conclusion of the antibody staining, the
cultures were mounted in a solution containing
1,4-diazobicyclo[2,2,2]octane to inhibit bleaching of the
fluorescence (Johnson et al., 1982).
In the present paper we have analyzed several properties of the
CA+ cells that develop in neural crest cultures in light of the
known properties of CA+ neural crest-derived cells in vivo. We
wished to determine in some detail the extent and nature of the
differentiation of the CA+ cells that develop in our neural crest
cultures. Our results suggest that the CA+ cells that de- velop in
our neural crest cultures possess a phenotype inter- mediate
between that of principal sympathetic neurons and ad- renal
chromaffin cells, and resembles in many respects SIF cells that are
found in autonomic ganglia and extra-adrenal chro- maffin tissue of
many species. The data presented here provide a more detailed
description of the phenotypic properties of the CA+ cells that
arise in neural crest cultures than has been avail- able
previously. This information should prove useful in our analysis of
how activation of the CA phenotype occurs in some neural crest
cells, but not in others, during embryogenesis.
Portions ofthis work were presented at the 16th Annual Meet- ing
of the Society for Neuroscience (Maxwell and Christie, 1986).
Uptake experiments. Growth medium was removed from the cells and
the cultures were preincubated for 10 min in 0.65 ml of medium
containing Dulbecco’s modified Eagle’s medium with 4.5 gm/liter
glu- cose without serum or chick embryo extract but with 1 mM
ascorbate, 0.1 mM pargyline, and test drug where indicated.
Cultures were then incubated for 30 min at 36.K in 5% CO, in the
same medium con- taining 0.5 PM 3H-norepinephrine (NE). At the
conclusion of the in- cubation, the cells were washed 5 times in
HBSS-HEPES containing 1 mM nonradioactive NE, scraped from the dish
into 100 ~1 HBSS-HEPES, and solubilized with 0.5 ml Protosol (New
England Nuclear). Samples were then treated with 100 ~1 of H,O, at
60°C for 30 min to bleach out any pigmentation resulting from
melanocytes present in the sample. The radioactivity in the sample
was measured in a liquid-scintillation spectrometer using
Econofluor cocktail (New England Nuclear).
6-Hvdroxvdooamine treatment. For treatment with 6-hvdroxvdo-
pamink, a 160 i stock solution of 8 mg/ml was prepared freshin
H&S- HEPES. At the time of treatment, fresh ascorbate at 20
&ml was also Materials and Methods
Neural crest cultures. Cultures of neural crest cells were
prepared from added to the culture to retard oxidation of the
6-hydroxydopamine. the trunk region of embryonic day 2, stage 13
(Zacchei, 196 1) Japanese Reserpine treatment. Reserpine was
prepared as a 1000 x stock so- quail embryos (Coturnix coturnix)
and grown as described previously lution of 0.1 M in dimethyl
sulfoxide. Control cultures received an equal (Maxwell et al.,
1982) with the minor modifications described below. volume of
dimethyl sulfoxide without reserpine. Briefly, neural tubes
containing the neural crest were isolated from sur- Electron
microscopy. Preparation of cultured cells for electron mi- rounding
tissues and were plated and grown in a medium containing croscopy
was carried out with the cells in situ in the culture dish.
Cultures 37.5 ml Dulbecco’s modified Eagle’s medium with 4.5
gm/liter glucose, DH 7.4 (Gibco). 37.5 ml F-12. OH 7.4 (Gibco). 15
ml horse serum
-
The Journal of Neuroscience, November 1987, 7(11) 3751
2ooo- -CA+
DAYS IN CULTURE Figure 1. Developmental appearance of TH+ and
CA+ cells in neural crest cultures. Cultures were grown as
described in Materials and Meth- ods and sister cultures were
processed to reveal either TH+ or CA+ cells on the days indicated.
The number of cells/culture positive for each trait was scored.
Values are presented as the means & SEM. Note that the ordinate
is plotted on a log scale.
The cultures were then stained en bloc using 1.0% uranyl acetate
in 0.05 M sodium acetate, pH 5.0, for 2 hr in the dark at 4°C.
Cultures were then rinsed 3 times in sodium acetate buffer,
dehydrated in a graded series of ethanols, and embedded in Polybed
8 12 resin (Polysciences). They were then cured at 60°C for 24 hr.
Marked areas of the embedded cultures were cut out with a jeweler’s
saw, mounted on epoxy stubs using cyanoacrylate cement, and trimmed
to the appropriate size for thin-sectioning.
Sections 80-90 nm thick were cut on a Sorvall Porter-Blum MT-2B
ultramicrotome and collected on 150-200-mesh copper grids. Most
cultures were sectioned in a plane parallel to the bottom of the
dish, starting in a plane above the cells so as not to miss any
cells of interest. Thin sections were collected at 1 pm intervals,
beginning as soon as cells were encountered in the block, until the
substrate was reached. Some cultures were sectioned perpendicular
to the bottom of the dish. Sections were grid-stained in 5% uranyl
acetate at 36°C for 30 min and then by 0.8% lead citrate at room
temperature for 5 min.
Sections were analyzed on a Phillips 300 electron microscope.
Ran- dom nonduplicated fields with GV-containing cells were
photographed by systematically scanning a single thin section and
photographing each field encountered until 14 exposures were
completed or all the fields with GV had been photographed. Only a
single thin section from a given field containing CA+ cells was
analyzed for morphometry, thus elim- inating the possibility that
any GV were sampled more than once. Mi- crographs used for
determinations of GV diameters were photographed at 50,000 x
magnification and subsequently printed at a final magnifi- cation
of 125,000 x . Micrographs used for vesicle-frequency determi-
nations were photographed at 4000x magnification and printed at
12,000 x . A calibration grid was photographed with each sample in
order to record the exact magnification, and the microscope
objective lens current was reduced to 0 before each micrograph was
taken to reduce the effects of hysteresis.
For morphometry, the outines of GV, cells, or nuclear profiles
were digitized on a digitizing tablet (Numonics) interfaced to an
LSI-11 com- puter (Digital Equipment Corporation). Using perimeter
points, the area and mean diameter of the profiles were calculated
as described in detail by Oliver (1985). Briefly, the longest axis
of a given profile was defined as the longest line that divided the
area into 2 equal areas. The average diameter was the average of
twice the radial distance from the long axis to each perimeter
point. At 125,000 x magnification, the lower limit of resolution of
vesicle-diameter measurements using this instrumentation was 4 nm.
Since the observed GV diameters differed from the true mean
diameters. because of the sectioning of some GV through planes
other than the equator, the corrected mean GV diameter was
calculated using the method of Giger and Riedwyl, as described by
Weibel(1979).
Vesicle frequency was calculated by counting the number of GV in
a given cell profile and then dividing that number by the area
ofcytoplasm, as expressed in a square microns. In cases where a
nuclear profile was present in a cell profile, the nuclear area was
subtracted from the total cellular area to yield the cytoplasmic
area.
Figure 2. Morphology of TH+ cells that develop in neural crest
cul- tures. The photograph shows TH+ cells in a neural crest
culture after 7 d in vitro, visualized using using antiserum to TH
and indirect im- munofluorescence, as described in Materials and
Methods. Immuno- reactivity to TH is observed in cell bodies and in
cellular processes. The morphology of the TH+ cells is very similar
to that previously reported for CA+ cells. Magnification, x
592.
Results Immunocytochemistry and pharmacology of CA+ cells
Previous studies have documented some of the features of the
histochemically and biochemically detectable CA+ cells that
differentiate in neural crest organ and cell cultures (Cohen, 1972,
1977; Norr, 1973; Kahn et al., 1980; Sieber-Blum and Cohen, 1980;
Fauquet et al., 1981; Maxwell et al., 1982; Kahn and Sieber-Blum,
1983; Maxwell and Sietz, 1983, 1985; Howard and Bronner-Fraser,
1986). We wished to examine several as- pects of the adrenergic
phenotype of these CA+ cells in more detail. As one part ofthis
analysis, we have compared the pattern of appearance of TH+ and CA+
cells in these cultures. In particular, we wished to determine
whether TH+ cells appeared prior to cells with endogenous CA
stores. We also wished to determine whether significant numbers of
cells were TH+, but did not contain detectable CA. As shown in
Figure 1, neither TH+ nor CA+ cells were observed prior to 4 d in
vitro under the cell culture conditions used. By 4 d in vitro,
small numbers of both cell types were seen. The number of both cell
types increased rapidly between 4 and 7 d in vitro. Similar numbers
of TH+ and CA+ cells were observed at any given time point. In
addition, the morphology of the TH+ cells in the cultures was very
similar to that observed for CA+ cells (Fig. 2). The TH+ cells were
often located in the region of the neural crest outgrowth adjacent
to the neural tube, as was the case for CA+ cells.
One diagnostic feature of neurons is the presence of neuro-
filaments in their cell bodies and processes (Weber et al., 1983).
Neural crest primary cultures after 7 d in vitro were processed to
visualize CA+ cells and then stained with antisera specific for
either the NF70 or the NF160 subunits of chicken neuro- filament
protein. We found that neither the processes nor the cell bodies of
CA+ cells were stained using either neurofilament antiserum (Fig.
3).
Neurofilament-immunoreactive processes were present in the
-
3752 Christie et al. - Adrenergic Neural Crest
Differentiation
Figure 3. The CA+ cells that develop in neural crest cultures do
not exhibit immunoreactivity to the NF70 neurofilament protein.
Neural crest cultures were grown for 7 d in vitro and then
processed in double-label experiments to reveal CA+ and NF70. A,
CA+ cells. B, The same field as in A viewed to reveal NF70+
structures. NF70 is present in cellular nrocesses that originate in
cells of the neural tube, but is absent from the CA+ cells.
Magnification, x 520.
cultures and were observed to originate mainly from cell bodies
present in the neural tube. To control for the possibility that the
histochemical procedure used to visualize CA+ cells may have
reduced the ability of neurofilament proteins to be rec- ognized by
the antibody, we compared neural crest cultures stained for
neurofilament using a standard immunocytochem- ical protocol, which
employed 4% paraformaldehyde fixation, to cultures stained after
visualization of CA+ cells. The pattern of neurofilament staining
was similar in both conditions. There was, however, some reduction
in the intensity of the neurofila- ment staining when CA+ cells
were also visualized.
As an alternative approach, we performed double-label im-
munocytochemical experiments using monoclononal antibodies directed
against TH produced in a mouse and the antisera di- rected against
the NF70 or NF160 chicken neurofilament pro- teins raised in
rabbits. As illustrated in Figure 4, these experi- ments
demonstrated that the TH+ cells did not contain
immunocytochemically detectable NF70 or NF160. The ab- sence of
NF70 and NF160 in TH+ cells at 7 d in vitro could have been
phenomenon of developmental maturation. For this reason, cultures
were also studied after longer times in vitro. In TH NF160
double-label experiments performed on cultures af- ter 11 d in
vitro, only 1 TH+ cell of 578 was NF 160+, while at 18 d in vitro
only 2 TH+ cells of 379 were observed to be NF160+. In TH NF70
double-label experiments performed after 19 and 26 d in vitro, none
of 267 TH+ cells scored was NF70+.
There was, however, abundant intense immunoreactivity to these
neurofilament proteins present in other cells in the cul- tures,
sometimes located in the same microscopic fields as the TH+ cells
(Fig. 4, 8, D). This immunoreactivity was present in numerous
processes whose cell bodies were located in the neural tube, which
had been intentionally retained in these cul- tures.
In addition to the NF70+ and NF160+ processes present in some
TH- cells, there were also some NF160+ and NF70+ cell bodies that
were TH- present in the neural crest outgrowth regions of the
culture (Fig. 5). The morphology of some of these NF+TH- cells
resembled that of neurons with an enlarged cell body and prominent
processes, while other NF+TH- cells had
a somewhat less neuronal morphology. The presence of these
NF+TH- cells suggests that neuronal cells that are not adre- nergic
differentiate in these neural crest cultures in addition to the TH+
cell population.
In other systems, neurons and neuroendocrine cells that me-
tabolize CAs have been shown to possess specific uptake systems for
these compounds (Iversen, 1967). In an effort to assess the uptake
properties of the CA+ cells that develop in neural crest cultures,
we have examined the sensitivity of this process to conditions with
known effects on CA uptake in other systems. As shown in Figure 6,
the uptake and storage of 3H-NE was blocked in the presence of
desmethylimipramine, an inhibitor of uptake I, and by incubation at
0°C. In contrast, this uptake and storage were unaffected by
normetanephrine, an inhibitor of uptake II.
Acute overnight application of the alkaloid reserpine, which
depletes CA stores from vesicles, resulted in the absence of
histochemically detectable CA+ cells from all neural crest cul-
tures (n = 6). Control experiments in which reserpine was added 10
min prior to CA histochemistry demonstrated that reserpine does not
interfere with the histochemical reaction used to detect CA+ cells.
The effect of acute overnight treatment with reser- pine was not
reversible over a period of at least 7 d. Chronic administration of
reserpine from day 2 to 7 in vitro resulted in the absence of CA+
cells in the cultures, but the presence of normal numbers of TH+
and SLI+ cells (Fig. 7). Since CA+ cells are not detectable until
day 4 in vitro at the earliest, this suggests that CA content in
vesicles is not required for the de- velopment of TH+ and SLI+
cells.
The compound 6-hydroxydopamine has profound effects on some
adrenergic neuron populations in the peripheral nervous system
(Thoenen and Tranzer, 1968; Tranzer and Thoenen, 1968). In adult
animals it destroys adrenergic nerve endings, while in neonatal
mammals the cell bodies are also destroyed (Angeletti and
Levi-Montalcini, 1970; Finch et al., 1973). When 6-hydroxydopamine
was administered to neural crest cultures that contained CA+ cells
on days 7 and 9 in vitro, and was then processed to reveal CA+
cells on day 11 in vitro, 1927 f 522 (mean + SEM; n = 4) CA+ cells
were present in the control condition and 1757 + 454 (n = 4) in the
6-hydroxydopamine-
-
The Journal of Neuroscience, November 1987, 7(11) 3753
Figure 4. Neurofilament proteins NF70 and NFl60 are not
detectable in TH+ cells that differentiate in neural crest
cultures. Neural crest cultures were grown for 7 d in vitro and
then processed in double-label indirect immunofluorescence
experiments, as described in Materials and Methods, to reveal TH+
cells and NF70+ structures or TH+ cells and NF160+ structures. A,
TH+ cells. B, NF70+ structures in the same field as in A. C, TH+
cells. D, NFl60+ structures in the same field as in C. In both the
NF70 and NFl60 cases, neurofilament-positive fibers whose cell
bodies of origin are in the neural tube are present, but no
neurofilament immunoreactivity is observed in the TH+ cells. In D,
the faint fluorescence seen in the region of the TH + cells is
equivalent to the background observed when TH + cells are stained
using the appropriate TRITC-labeled second antibodies and viewed
using FITC optics. Magnification, x 520.
treated cultures. The light-microscopic appearance of the cells
was the same in both conditions. Cultures that were allowed to grow
for up to 7 d after 6-hydroxydopamine treatment also showed no
effect on CA+ cell survival or light-microscopic morphology. In
contrast to this lack of effect on neural crest cells, the same
regimen of 6-hydroxydopamine was effective in the elimination of
brightly fluorescent CA+ cells from embry- onic day 11 quail
sympathetic ganglia grown in tissue culture.
NGF is a compound with well-documented effects on several
populations of neural crest-derived cells (Levi-Montalcini and
Angeletti, 1968; Greene and Shooter, 1980; Thoenen and Barde,
1980). Under growth conditions in which chick embryo extract was
present in the medium, the addition of exogenous NGF did not affect
the number of histochemically detectable CA+ cells that developed.
When cells were grown from day 2 to 7 in vitro in 0, 10, 50, and
100 rig/ml of NGF, there were 2528 + 409, 2186 rt: 413, 2107 -+
622, and 2327 f 566 CA+ cells present per culture, respectively, on
day 7 in vitro (mean + SEM; n = 5 in all cases). In addition, the
morphology of the CA+ cells was similar in the presence and absence
of exogenous NGF. Analysis of our growth medium with embryo extract
determined that there was less than 2 rig/ml NGF-like
immunoreactive material, as determined by a modified, one-site
radioimmu-
noassay (RIA) for NGF using antiserum to mouse NGF (M.
Rosenberg, personal communication). While chick embryo ex- tract
does not contain molecules that are immunologically sim- ilar to
mouse NGF, there may be other molecules present that may have NGF
activity (Cohen and Konigsberg, 1975; Gotz et al., 1986; Shelton et
al., 1986) and whose presence may prevent detection of a response
to exogenously added NGF. To try to rule out this possibility, a
second series of experiments was performed. In this set of
experiments, the cultures were grown in medium with chick embryo
extract, but no NGF, for 7 d in order to generate CA+ cells, and
then shifted into medium without chick embryo extract, but
containing 15% horse serum (which does not possess NGF activity),
with or without exog- enous NGF, at 100 r&ml. Six days after
the medium shift, the cultures were processed to reveal CA+ cells.
The results of these experiments showed that 1216 & 383 (mean +
SEM; n = 4) CA cells per culture were present without exogenous
NGF, while 1025 f 361 (n = 4) cells were present when exogenous NGF
was added. The morphology of the CA+ cells was the same in both
cases, with brightly fluorescent cell bodies and some pro- cesses
present. Thus, even in the absence of chick embryo ex- tract, NGF
did not alter the number of CA+ cells or their morphology.
-
3754 Chnstie et al. l Adrenerglc Neural Crest
Differentiation
Figure 5. Cells that are NF70+ and NF160+ but TH- are present in
the neural crest outgrowths. Neural crest cultures were grown for 7
d in vitro and then processed for double-label indirect
immunofluorescence to detect Th and NF70 or NF160. A, TH. B, NF70+
cell in the same field as in A. C, TH. D, The same field as in C,
viewed to reveal NF160+ structures. Magnification, x 520.
Ultrastructural properties of CA-t cells
While the light-microscopic data on the properties of CA+ cells
in neural crest cultures are informative in many respects, there
are certain properties that can only be analyzed at the ultra-
structural level. In particular, this is the case for information
concerning the vesicular content of the CA+ cells, which is one
important diagnostic feature for the type of CA+ cell present.
Accordingly, we have performed a correlative light- and elec-
tron-microscopic study of the ultrastructure of CA+ cells in neural
crest cultures at several stages of development.
The CA+ cells were a small subpopulation of the cells in our
neural crest cultures, and it was difficult to predict precisely
where in a culture the CA+ cells were present. This precluded a
random blind sampling of the cultures for any meaningful
ultrastructural analysis of CA+ cells. To overcome this diffi-
culty, we have used the method of Grill0 et al. (1974), which
allowed the identification of CA+ cells in the light microscope and
subsequent analysis of the same cells in the electron mi- croscope.
Using this procedure, we were able to identify CA+ cells that were
comparable in number, morphology, and spatial distribution in the
cultures to those observed using the method of Furness et al.
(1977).
When we examined areas of the cultures that contained CA+ cells
with the electron microscope, we observed, in addition to
melanocytes, large numbers of unpigmented cells. Many of these
unpigmented cells contained in their cytoplasms membrane- bound
vesicles with electron-dense cores, characteristic of CA- storage
vesicles observed in other systems (Figs. 8, 9). Our ex- amination
of thin sections collected at intervals from the top to the bottom
of the cultures showed that the GV-containing cells were
consistently found in the lower layers of the cultures ad- jacent
to the collagen substrate, where they were often beneath pigmented
or nonpigmented cells. Our analysis determined that the
GV-containing cells corresponded to CA+ cells seen in the light
microscope. We established that there was a one-to-one
correspondence between the number and spatial arrangement of CA+
and GV-containing cells (Fig. 10). Analysis of micro- scopic fields
without CA+ cells revealed no GV-containing cells.
The GV-containing cells were generally small, with cell bodies 1
O-20 pm along their longest axis. The nuclei were irregular in
shape and euchromatic. In many cases a prominent nucleolus was
present, in addition to characteristic small amounts of het-
erochromatin found in quail nuclei located adjacent to the nu-
clear membrane. The cytoplasm contained numerous mito- chondria and
Golgi profiles, a few areas of rough endoplasmic reticulum, and
abundant polyribosomes (Fig. 11). On average, the cytoplasmic
ground substance of the GV-containing cells was darker than that of
unpigmented cells that did not contain GV (Fig. 8). The principal
distinguishing feature of these cells,
-
The Journal of Neuroscience, November 1987, 7(11) 3755
0% 100 -
20”
WZ
80-
Y, 2: 60-
5: a w- 40-
z
& In 20-
Control DMI Normet. 0°C
Figure 6. Catecholamine uptake and storage properties of neural
crest cultures. Neural crest cultures after 7 d in vitro were
incubated with 0.5 FM ‘H-NE for 30 min and then processed as
described in Materials and Methods. The amount of cell-associated
radioactivity, accumulated un- der control conditions, in the
presence of 2 FM desmethylimipramine @MI), 10 PM normetanephrine
(Normet), or incubation at O’C, is shown. Values are expressed as a
percentage of the mean control value, which was normalized to 100%.
Error bars show * SEM with 5-7 determi- nations per condition.
however, was the presence of numerous GV. These were round or
oval, bounded by a typical unit membrane, and possessed an
electron-dense core with an electron-lucent halo. Typically the
electron-dense cores were located eccentrically in the GV.
Qualitatively, the GV-containing cells were morphologically
similar at all ages examined, as described above. Cells with GV in
their cytoplasm were detected as early as 4 d in vitro, but the
number of such cells was very small. Since the presence of the GV
was the most striking feature of the CA+ cells, we performed a
morphometric analysis on the GV present in cultures after 5, 7, 14,
and 2 1 d in vitro in order to determine quantitatively the
characteristics of these cells and the changes that occurred in
culture. As shown in Figure 12, the observed vesicle diameters
ranged from 50 to 600 nm, with a major fraction of vesicle
diameters clustered f 50 nm around a modal value in the range of
85-l 15 nm for the ages examined. There was no evidence of a
dramatic change in the pattern of vesicle diameters over the period
from day 5 to 21 in vitro (Fig. 12, Table l), although the
population of GV greater than 200 nm in diameter was some- what
greater at 21 d in vitro. The frequency of GV (number of GV/pm2 of
cytoplasm) also remained quite constant over the period of 5-21 d
in vitro. At 5, 7, 14, and 21 d in vitro, there were 2.55 f 2.60
(mean + SD) (n = 39), 3.19 + 2.66 (n = 89), 2.29 -+ 1.69 (n = 77)
and 2.06 -t 0.89 (n = 41) GV/hm2 of cytoplasm, respectively. The n
in this case refers to the number of cell profiles examined. On
average, the GV occupied ap- proximately 5% of the total
cytoplasmic area of the cell.
Fluorescent processes were seen extending from many CA+ cells
and were often observed to run among clusters of GV- containing
cells. These processes were characterized by a pale cytoplasm
lacking ribosomes, but containing numerous micro- tubules arrayed
longitudinally with dense core GV often lined up along the
microtubules (Fig. 13).
In cultures after 5 and 7 d in vitro, synaptic profiles were not
observed. However, in cultures after 14 and 2 1 d in vitro,
nerve
” I CA I TH I SLI u x 40- 40- 0-
f$ 30- 30- 6- 5! 3 *o- 20- 4- z j IO- IO- 2-
8 0, -+ -+ -+
Figure 7. The effect of chronic reserpine treatment on the
appearance of CA+, TH+, and SLI+ cells. Neural crest cultures were
grown with either 10 1~ reserpine (+) or dimethyl sulfoxide (the
reserpine solvent, -) for days 2-7 in vitro, and then analyzed on
day 7 for the number of CA+. TH+. or SLI+ cells. as described in
Materials and Methods. Values are expressed as the mean f SEM with
5-7 determinations for each condition.
terminals were frequently seen adjacent to GV-containing cell
bodies. In some terminal profiles, synaptic membrane special-
izations were evident (Fig. 14). One typical terminal contained
predominantly clear vesicles, 59 +- 10 nm in diameter (mean + SD; n
= 37 vesicles), with a few dense core vesicles 104 +- 26 nm in
diameter (mean f SD; n = 25 vesicles). The cell bodies of origin of
these terminals are not known.
Discussion In the present paper we have examined several
properties of the CA+ cells that differentiate in neural crest
cultures in order to assess the nature and extent of their
development. These experiments, together with other available
information, have allowed us to conclude that the CA+ cells that
develop in our neural crest cultures resemble SIF cells in many
respects. This phenotype is consistent with these CA+ cells being
either a SIF-like developmental intermediate that is stabilized by
our culture conditions or a terminally differentiated SIF-like cell
type such as those found in some adult sympathetic ganglia and
extra-adrenal chromaffin tissue.
Correspondence of TH+ and CA+ cells The CA+ cells in these
neural crest cultures develop from a CA- cell population that is
also TH - . The temporal appear- ances of TH+ and CA+ cells are
very similar, suggesting that, at least with a time resolution of
days, there is no appreciable lag between the appearance of TH and
the capabiity to synthe- size and store CA. These findings are in
agreement with previous biochemical observations comparing the time
courses of ap- pearance of dopamine+hydroxylase enzyme activity and
the production and storage of radiolabeled CA from radioactive
tyrosine by intact cells (Kahn et al., 1980; Maxwell et al., 1982).
The present results, comparing the number of TH+ and CA+ cells as a
function of time in culture, also demonstrate that, after 7 d in
vitro, the number of cells with both markers in their cell bodies
is similar. This suggests that there is no large population of
cells that is TH+ but CA- in the cell body. In addition, there does
not appear to be a large population of CA+ processes present
without CA+ cell bodies. This phenotype might be expected if large
numbers of cells resembling principal sym-
-
3756 Christie et al. - Adrenergic Neural Crest
Differentiation
Figure 8. Ultrastructure of CA+ cell cytoplasm at 5 d in vitro.
Electron micrograph showing portions of 2 cells from a neural crest
primary culture after 5 d in vitro. The cell labeled G contains
numerous GV ranging from 50 to 600 nm (see Fig. 12). These GV
contain eccentrically located dense cores. Also present in the
cytoplasm are numerous mitochondria and free ribosomes. The
cytoplasm of the cells with GV is darker than that of the adjacent
unpigmented cell (U), which lacks GV. Magnification, x 14,560.
pathetic neurons were present in these cultures (Doupe et al.,
1985a, b).
Neurojlament immunoreactivity One very characteristic feature of
neuronal differentiation is the expression of neurofilament
proteins. The expression of neu- rofilament proteins is a
relatively early event in the differentia- tion of many neuron
populations, including some autonomic neurons (Tapscott et al., 198
1; Jacobs et al., 1982; Ziller et al., 1983; Cochard and Paulin,
1984; Anderson and Axel, 1986). Our data indicate that virtually
all the CA+ and TH+ cells that differentiate in our neural crest
cultures lack detectable NF70 and NF160 in either their cell bodies
or processes. This lack of neurofilament immunoreactivity appears
to be a relatively sta- ble trait of the CA+ and TH+ cells in these
cultures, since in cultures kept as long as 26 d in vitro the TH+
cells did not possess neurofilament immunoreactivity.
It has been reported that in PC1 2 cells grown in the absence of
NGF, and in embryonic rat neuronal precursors, neurofila- ment
immunoreactivity can be detected in a perinuclear “ball” (Lee and
Page, 1984; Anderson and Axel, 1986). Also, immu- noreactivity to
NF160 has been observed in young avian neural crest cells in vivo
that are likely the precursors ofenteric neurons (Payette et al.,
1984). We have not observed such neurofilament staining in our CA+
and TH+ cells in neural crest cultures, which suggests that our CA+
and TH+ cells are at a different stage of developmental maturation
than these other cell types.
Many fibers and some cell bodies that do have NF70 and NF160 can
be detected in other cells in these cultures. This
neurofilament immunoreactivity in CA - and TH - cells serves as
an internal control for the sensitivity and specificity of our
immunocytochemical procedures. In addition, the presence of
neurofilament-positive TH- cells in the neural crest outgrowths
points to the differentiation of neuronal cell types in these trunk
neural crest cultures that do not have adrenergic traits. These may
include cells with some of the properties of sensory, para-
sympathetic, and enteric neurons, as well as nonadrenergic sym-
pathetic neurons (Kahn et al., 1980; Fauquet et al., 198 1; Max-
well et al., 1982; Ziller et al., 1983; Mackey et al., 1986;
Sieber- Blum et al., 1986).
CA uptake and storage properties Our data on the uptake and
storage of )H-NE are consistent with the view that some cells in
our neural crest cultures acquire an uptake I-like system similar
to that observed in other de- veloping and mature adrenergic cell
populations in vivo and in vitro (Iversen, 1967; Patterson et al.,
1975; Greene and Rein, 1977; Rothman et al., 1978). Previous
autoradiographic evi- dence and histochemistry for CA+ cells, when
correlated with data on the evoked release of 3H-NE, strongly
suggested that it is the subpopulation of histochemically
detectable CA+ cells that is responsible for the uptake and storage
of 3H-NE (Maxwell and Sietz, 1983). This CA+ cell population is
sensitive to re-
‘set-pine administered in both acute and chronic regimens. It
has been reported that some interneuron populations of older em-
bryonic sympathetic ganglia are reserpine-resistant (Benitez et
al., 1973, 1974). This does not appear to be the case for the CA+
cells that develop in our cultures. While the chronic
-
The Journal of Neuroscience, November 1997, 7(11) 3757
Figure 9. Comparison of the ultrastructure of CA+, unpigmented
and melanocyte cytoplasm at 5 d in vitro. Electron micrograph of
portions of 3 cells from a neural cresf cultures after 5 d in
vitro. At the top is a pigmented cell (P) containing melanosomes.
At the left is a portion of a CA+ cell (G) containing numerous GV.
At the bottom is an unpigmented cell (U) without GV. Magnification,
x 12,207.
administration of reserpine eliminates the appearance of CA+
cells, TH + and SLI + cells are present in normal numbers. These
data indicate that reserpine acts on these cells by depletion of CA
stores and not by killing CA+ cells. Furthermore, these data show
that the presence of CA stores in vesicles is not necessary for the
development of SLI. Since reserpine is present from day 2 to 7 in
these experiments, and CA is not detectable until day 4 in control
cultures, the SLI+ cells must arise in the absence of
histochemically detectable CA. Since we assayed SLI cell number and
not SLI content, we do not know if reserpine altered peptide
content. Other studies have demonstrated varied effects of
reserpine on peptide metabolism, which are probably depen- dent on
the cell type and species examined (Wilson et al., 1980, 1981; Bohn
et al., 1983; Kessler, 1985).
The CA+ cells that develop in neural crest cultures do not
exhibit a dramatic response to exogenous NGF with respect to either
the number of cells that differentiates and survives or the
light-microscopic morphology of the cells. This is in contrast to
the very strong dependence on NGF for survival exhibited by older
embryonic avian sympathetic and sensory neurons (Greene, 1977a, b).
However, during the early stages of avian sympathetic and sensory
ganglion development, neuronal survival and dif- ferentiation do
not require the presence of NGF (Winick and Greenberg, 1965;
Partlow and Larrabee, 1971; Barde et al., 1980; Edgar et al.,
1981). Also, the SIF-like cells that develop in tissue cultures of
older embryonic chick sympathetic ganglia do not require NGF for
their survival and differentiation (Ja- cobowitz and Greene, 1974).
Thus, the apparent lack of an NGF response by the CA+ cells that
differentiate in our neural crest cultures is consistent with
either a SIF-like cell phenotype or an
early stage of principal sympathetic neuron development. Our
data are consistent with recent reports that CA+ and
THS cells in neural crest cultures grown in the absence of NGF
do not possess a significant ability to bind NGF, suggesting that
they lack NGF receptors (Bemd, 1986a; Greiner and Guroff, 1986).
Thus, the most straightforward explanation for the lack of a
response to NGF by CA+ cells is that they lack NGF receptors. It
should be noted, however, that other cells in these neural crest
cultures do possess NGF receptors (End et al., 1983; Bemd, 1985,
1986b; Greiner et al., 1985, 1986). Recent data by Bemd (1986a)
suggest that when neural crest cultures are grown in the presence
of NGF, some TH+ cells with NGF receptors are present. The number
of these NGF-binding TH+ cells has not been reported. Unless the
number is quite large, it is possible that they would not have been
detected by us given the variability of total CA+ cell number from
culture to culture. In addition, we cannot exclude the possibility
that CA+ cells might exhibit an NGF response by an alteration of
metabolism without affecting cell survival and hence cell
number.
The CA+ cells in neural crest cultures are resistant to the
administration ofdoses of 6-hydroxydopamine that are effective in
killing embryonic quail sympathetic neurons. This resistance to
6-hydroxydopamine treatment may be related to the relative
insensitivity of these cultures to NGF. Aloe et al. (1975) have
shown that NGF can protect rat sympathetic neurons against the
effects of 6-hydroxydopamine with respect to cell survival. They
suggested that perhaps cells that are insensitive to NGF may also
be resistant to 6-hydroxydopamine (Levi-Montalcini and Aloe, 1980).
While our CA+ cells in neural crest cultures are not killed by
6-hydroxydopamine, it is possible that the
-
3758 Christle et al. - Adrenergic Neural Crest
Differentiation
Figure IO. Correspondence of CA+ cells observed in the light
microscope and GV-containing cells seen in the electron microscope.
The inset shows 3 CA+ cells in a neural crest culture after 2 1 d
in vitro. The electron micrograph shows that the same 3 cells
contain abundant GV in their cytoplasm. The GV are absent from
surrounding cells that are not CA+. Arrows indicate the same
relative position in both photographs. Magnification, light
micrograph, x 475; electron m
-
The Journal of Neuroscience, November 1987, 7(11) 3759
Figure II. Ultrastructure of CA+ cells after 14 d in vitro.
Electron micrograph showing portions of 3 cells from a neural crest
culture after 14 d in vitro, containing GV. The cell at the lower
right also contains several Golgi figures, numerous mitochondria,
rough endoplasmic reticulum, and free ribosomes. Magnification, x
11,250.
Rechardt, 1974; Hervonen and Eranko, 1975). These cells per-
sist in cultured ganglia in the absence of NGF. The identity of
these cells has been determined differently by different authors.
Lever and Presley (197 1) suggested that they are analagous to
“interneurons” (SIF cells) of the rat sympathetic ganglia (Mat-
thews and Raisman, 1969). However, Benitez et al. (1973, 1974)
concluded that many of these cells are actually young principal
neurons because of their light-microscopic morphology. Her- vonen
and Eranko (1975) contended that they are sympathi- coblasts and
not SIF cells on the basis of the arrangement of the chromatin in
the nucleus. Our observed modal GV diam- eters were similar to
those of the vesicles present in the cyto- plasm of adrenergic
cells found in sclerotomal cultures initiated from 3-d-old quail
embryos (Smith and Fauquet, 1984). Precise size comparisons are not
possible because of differences in fix- ation and staining
methods.
On the basis of our analysis, it seems that the CA+ cells
that
differentiate in our quail neural crest cultures are morphologi-
cally similar to the cells with CA storage granules in their cy-
toplasm found in avian sympathetic ganglia in tissue culture and
transiently during embryogenesis in vivo. They may be sim- ilar in
phenotype to the first CA+ cells that are observed to develop in
vivo as the primary sympathetic chains form (Enemar et al., 1965;
Wechsler and Schmekel, 1967). In many respects these CA+ cells
resemble type I SIF cells. These cells with CA+ fluorescence in
their cell body in neural crest cultures possess abundant GV with
modal diameters near 100 nm. On the basis of the unimodal nature of
the histograms of GV diameter, it seems likely that we are dealing
with a relatively homogeneous class of cells, at least with respect
to their ultrastructural prop- erties. The CA+ cells in our
cultures remain quite constant with regard to the size and density
of GV present in the cytoplasm over a period of 3 weeks in vitro.
As noted above, these CA+ cells also do not exhibit neurofilament
immunoreactivity. In
Table 1. Ultrastructural properties of CA+ cells
Corrected GV with Total Observed mean Observed observed no.
of
Days in mean GV diam modal GV diameters Total no. of cells vitro
GV diam (nm) 6-d diam (nm) >200 nm (46) GV analyzed analyzed
5 145 * 56 172 105 19.4 500 15 7 137 * 74 145 85 12.4 410 28
14 160 I!Z 91 174 115 17.1 410 45 21 190 * 122 219 85 33.0 376
27
Observed mean GV diameter is expressed *SD for the total number
of GV analyzed. The corrected mean GV diameter was calculated by
the method of Giger and Riedwyl, as described in Weibel(1979).
-
3760 Christie et al. * Adrenergic Neural Crest
Differentiation
5 DAY
14 DAY
21 DAY
GRANULAR VESICLE DIAMETER (nm)
Figure 12. Histograms of GV diameters as a function of time in
cul- ture. Neural crest cultures were grown for 5, 7, 14, and 21 d
in vitro and then processed to reveal CA+ cells in the light
microscope, followed by electron microscopy and morphometry of GV
diameters, as described in Materials and Methods. The percentage of
GV with diameters that fell in a given bin 10 nm wide is shown.
contrast, the SIF-like cells observed in vivo probably either
de- generate or are transformed into principal sympathetic neurons
(Wechsler and Schmekel, 1967; Luckenbill-Edds and van Horn, 1980).
Such a transformation can also occur when SIF-like cells from rat
sympatheic ganglia or rat and chick adrenal chromaffin
Figure 13. Presence of GV in cellular processes. Electron
micrograph showing a segment of a neurite located in a cluster of
GV-containing cells in a neural crest culture after 14 d in vitro.
Large GV vesicles are aligned along an array of microtubules
(arrows). Magnification, x 90,750.
Figure 14. A nerve terminal contacting a GV-containing cell. A
ter- minal profile (T) in association with a granule-containing
cell (G) in a neural crest culture after 14 d in vitro. The
terminal contains mostly small (50 nm) agranular vesicles and some
larger (100 nm) granular vesicles, and rests in a depression in the
granule-containing cell surface. A thickening (arrow) can be seen
on the opposed membrane. Magni- fication, x21,150.
cells are grown under the appropriate conditions in tissue
culture (Unsicker et al., 1978; Doupe et al., 1985a, b; Anderson
and Axel, 1985, 1986; Shaw and Letoumeau, 1986).
CA+ cells in neural crest cultures and the adrenergic
developmental program
The available evidence, both at the light- and electron-micro-
scopic levels, enables us to conclude that the CA+ cells in our
neural crest cultures possess a phenotype that in many aspects
resembles that observed in SIF cells and that is intermediate
between the adrenal chromaffin cell and principal sympathetic
neuron phenotypes. This conclusion is based on several lines of
evidence. The presence of CA clearly establishes that they are more
differentiated than CA- progenitors of the neural crest. If these
CA+ cells were mature principal neurons, we would expect them to
have CA fluorescence in processes, but not in cell bodies, to
depend on NGF for survival and differentiation, to be neu-
rofilament-positive, and to possess small GV in endings but not in
the cell body. However, our findings of intense fluorescence in the
cell body, the lack of prominent NGF and 6-hydroxy- dopamine
responses, and the absence of neurofilament immu- noreactivity,
together with ultrastructural data, strongly argue against the CA+
cells as being differentiated principal neuron- like cells. Largely
on the basis of GV size and the presence of cellular processes,
these CA+ cells do not appear to be mature adrenal chromaffin
cells. This, then, suggests that the CA+ cells that develop in our
neural crest cultures either resemble SIF- like cells or a more
primitive, sympathoadrenal precursor.
The difference between SIF-like cells and the postulated, more
primitive, sympathoadrenal precursor cell is that the SIF cell is
responsive to NGF and glucocorticoids, while the sympathoad- renal
precursor is by definition not yet responsive to NGF or possibly to
glucocorticoids (Landis and Patterson, 198 1; Doupe et al., 1985b,
Anderson and Axel, 1986). In the model of Landis and Patterson (198
1) and Doupe et al. (1985b), we would expect SIF cells to be
responsive to both NGF and glucocorticoids, although in very
different ways. One would predict that in the
-
The Journal of Neuroscience, November 1987, 7(11) 3761
presence of NGF, cell bodies would lose their intense CA flu-
orescence and would acquire other neuronal traits, such as the
presence of neurofilaments and extension of long CA+ pro- cesses.
We have not systematically examined all of these traits in our
neural crest cultures, but we do not see a loss of cell body
fluorescence and the retention of a large number of CA flu-
orescent processes in the presence of exogenous NGF. One re-
servation is that the SIF cell conversions observed by Doupe et al.
(1985a, b) are quite slow, so that we simply may not have waited
long enough for the transition to occur. However, it should be
noted that the neuroendocrine precursor cell isolated by Anderson
and Axe1 ( 1986) can show NGF or glucocorticoid responses within
3-4 d. One possible explanation is that the CA+ cells in our
cultures are prevented from undergoing further developmental
changes because they lack exposure to some ap- propriate signal.
For example, the cells may require interaction with somitic
mesenchyme or target tissue to become competent to respond to NGF
(Smith and Fauquet, 1984; Davies et al., 1987). Another point is
that since our cells are grown in medium containing serum and
embryo extract, perhaps the presence of glucocorticoids in the
medium acts to mask any effect of ex- ogenously added NGF (Unsicker
et al., 1978). At present, we cannot distinguish among these
possibilities.
It is also possible that the CA+ cells that develop in our
neural crest cultures are neither a sympathoadrenal precursor nor a
SIF-like developmental intermediate, but are rather a
differentiated endpoint cell type, such as the CA+ cells found in
the carotid body and extra-adrenal chromaffin tissue. This point
will be resolved if conditions can be found that enable us to make
the CA+ cells in our neural crest cultures progress to another
point in development, or if we can find additional phe- notypic
markers that further clarify the developmental status of these
cells. It is also worth noting that there may be significant
species differences with respect to the precise regulation of de-
velopmental pathways in the cell lineage of the autonomic ner- vous
system (Edgar et al., 1981). Thus, not all aspects of the models
generated from experiments using rodents necessarily apply to avian
neuronal development.
Conclusions
In summary, our present light- and electron-microscopic data,
together with previous results obtained by ourselves and others,
indicate that the CA+ cells that develop in our neural crest
cultures resemble, in many respects, SIF-like cells. Under the
conditions of our cultures, the CA+ cell type seems quite stable
for at least 3 weeks in vitro. As such, these cells offer excellent
opportunities for future investigations concerning the mecha- nism
by which neural crest cells acquire adrenergic traits and the
specific cues necessary for them to execute additional steps in the
adrenergic pathway of differentiation.
References Aloe, L., E. Mugnaini, and R. Levi-Montalcini (1975)
Light and elec-
tron microscope studies on the excessive growth of sympathetic
gan- alia in rats iniected from birth with 60HDA and NGF. Arch.
Ital. Viol. 113: 326-353.
Anderson, D. J., and R. Axe1 (1985) Molecular probes for the
devel- opment and plasticity of neural crest derivatives. Cell 42:
649-662.
Anderson, D. J., and R. Axe1 (1986) A bipotential neuroendocrine
precursor whose choice of cell fate is determined by NGF and glu-
cocorticoids. Cell 47: 1079-1090.
Angeletti, P. U., and R. Levi-Montalcini (1970) Sympathetic
nerve cell destruction in newborn mammals by 6-hydroxydopamine.
Proc. Natl. Acad. Sci. USA 65: 114-121.
Barde, Y. A., D. Edgar, and H. Thoenen (1980) Sensory neurons in
culture: Changing requirements for survival factors during
embryonic development. Proc. Natl. Acad. Sci USA 77: 1199-1203.
Benitez, H. H., M. R. Murray, and L. J. Cote (1973) Responses of
sympathetic chain-ganglia isolated in organotypic culture to agents
affecting adrenergic neurons: Fluorescence histochemistry. Exp.
Neu- rol. 39: 424-448.
Benitez, H. H., E. B. Masurovsky, and M. R. Murray (1974) Inter-
neurons of the sympathetic ganglia in organotypic culture. A
sugges- tion as to their function based on three types of study. J.
Neurocytol. 3: 363-384.
Bennett, G. S., S. J. Tapscott, C. DiLullo, and H. Holtzer
(1984) Dif- ferential binding of antibodies against the
neurofilament triplet pro- teins in different avian neurons. Brain
Res. 304: 291-302.
Bennett, T., and T. Malmfors (1970) The adrenergic nervous
system of the the domestic fowl (Callus domesticus L.). Z.
Zcllforsch. Mi- krosk. Anat. 106: 22-50.
Bemd, P. (1985) Appearance of nerve growth factor receptors on
cul- tured neural crest cells. Dev. Biol. 112: 145-l 56.
Bemd, P. (1986a) Presence and/or absence of nerve growth factor
receptors on neuron-like cells in long term neural crest cultures.
Sot. Neurosci. Abstr. 12: 1094.
Bemd, P. (1986b) Characterization of nerve growth factor binding
to cultured neural crest cells: Evidence of an early developmental
form of the NGF receptor. Dev. Biol. 115: 4 15-424.
Black, I. B. (1982) Stages of neurotransmitter development in
auto- nomic neurons. Science 215: 1198-l 204.
Biscoe, T. J. (1971) Carotid body: Structure and function.
Physiol. Rev. 51: 437-495.
Bohn, M. C., J. A. Kessler, L. Golightly, and I. B. Black (1983)
Ap- pearance of enkephalin immunoreactivity in rat adrenal medulla
fol- lowing treatment with nicotinic antagonists or reserpine. Cell
Tissue Res. 231: 469-479.
Bumstock, G., and M. Costa (1975) Adrenergic Neurons, Wiley, New
York.
Cahn, R. D., H. G. Coon, and M. R. Cahn (1967) Cell culture and
cloning techniques. In Methods in Developmental Biology, F. W. Wilt
and N. K. Wessels, eds., pp. 493-530, Crowell, New York.
Charnley, J. H., G. E. Mark, G. R. Campbell, and G. Bumstock
(1972) Svmuathetic aanelia in culture I. Neurons. Z. Zellforsch.
Mikrosk. Anat. 135: 287-314.
Cochard, P., and D. Paulin (1984) Initial expression of
neurofilaments and vimentin in the central and peripheral nervous
system of the mouse embryo in vivo. J. Neurosci. 4: 2080-2094.
Cochard, P., M. Goldstein, and I. B. Black (1978) Ontogenetic
ap- pearance and disappearance of tyrosine hydroxylase and
catechol- amines in the rat embryo. Proc. Natl. Acad. Sci. USA 75:
2986-2990.
Cochard, P., M. Goldstein, and I. B. Black (1979.) Initial
development of the noradrenergic phenotype in autonomic neuroblasts
of the rat embryo in vivo. Dev. Biol. 71: 100-l 14.
Cohen, A. M. (1972) Factors directing the expression of
sympathetic nerve traits in cells of neural crest origin. J. Exp.
Zool. 917: 167-l 82.
Cohen, A. M. (1977) Independent expression of the adrenergic
phe- notype by neural crest cells in vitro. Proc. Natl. Acad. Sci.
USA 74: 2899-2903.
Cohen, A. M., and I. R. Konigsberg (1975) A clonal approach to
the problem of neural crest determination. Dev. Biol. 46:
262-280.
Coupland, R. E. (1965) The Natural History of the ChromaJin
Cell, Longman, New York.
Davies, A. M., C. Bandtlow, R. Heumann, S. Korsching, H. Rohrer,
and H. Thoenen (1987) Timing and site of nerve growth factor
synthesis in relation to innervation and expression of the
receptor. Nature 326: 353-358.
Doupe, A. J., and P. H. Patterson (1980) Glucocorticoids and the
developing nervous system. In Current Topics in Neuroendocrinology,
D. Ganten and D. Pfaff, eds., pp. 23-43, Springer-Verlag,
Berlin.
Doupe, A. J., S. C. Landis, and P. H. Patterson (1985a)
Environmental influences in the development of neural crest
derivatives: Glucocor- ticoids, growth factors, and chromaffin cell
plasticity. J. Neurosci. 5: 2119-2142.
Doupe, A. J., P. H. Patterson, and S. C. Landis (1985b) Small
intensely fluorescent cells in culture: Role ofglucocorticoids and
growth factors in their development and interconversions with other
neural crest derivatives. J. Neurosci. 5: 2 143-2160.
Edgar, D., Y. A. Barde, and H. Thoenen (1981) Subpopulations
of
-
3762 Christie et al. - Adrenergic Neural Crest
Differentiation
cultured chick sympathetic neurons differ in their requirements
for survival factors. Nature 289: 294-295.
Ehrmann, R. L., and G. 0. Gey (1956) The growth of cells on a
transparent gel of reconstituted rat-tail collagen. J. Natl. Cancer
Inst. 16: 1375-1403.
End, D., L. Pevzner, A. Lloyd, and G. Guroff (1983)
Identification of nerve growth factor receptors in primary cultures
of chick neural crest cells. Dev. Brain Res. 7: 131-136.
Enemar, A., B. Flack, and R. Hakanson (1965) Observations on the
appearance of norepinephrine in the sympathetic nervous system of
the chick embrvo. Dev. Biol. 11: 268-283.
Eranko, L. (1972) Ultrastructure of the developing sympathetic
nerve cell and the storage of catecholamines. Brainkes. 46:-l
59-172.
Eranko. 0. (ed.) (1975) Structure and function of the small.
intenselv fluorescent sympathetic cells. In Fogarty International
C&ter Pro: ceedings, United States Government Printing Office,
Washington, D.C.
Fauquet, M., J. Smith, C. Ziller, and N. M. LeDouarin (1981)
Dif- ferentiation of autonomic precursor cells in vitro:
Cholinergic and adrenergic traits in cultured neural crest cells.
J. Neurosci. 1: 478- 492.
Finch, L., G. .Hausler, and H. Thoenen (1973) A comparison of
the effects of chemical sympathectomy by 6-hydroxydopamine in new-
born and adult rats. Br. J. Pharmacol. 47: 249-260.
Fumess, J. B., M. Costa, and A. J. Wilson (1977) Water-stable
fluo- rophores, produced by reaction with aldehyde solutions, for
the his- tochemical localization of catechol- and indolethylamines.
Histo- chemistry 52: 159-l 70.
Gabella, G. (1976) Structure of the Autonomic Nervous System,
Chap- man and Hall, London.
Garcia-Arraras, J. E., M. Chanconie, and J. E. Fontaine-Perus
(1984) In vivo and in vitro development of somatostatin-like
immunoreac- tivity in the peripheral nervous system ofquail
embryos. J. Neurosci. 4: 1549-1558.
Gotz, R., R. Meier, M. Becker-Andre, R. Heumann, and H. Thoenen
(1986) Molecular cloning of bovine and chick nerve growth factor:
Delineation of conserved and unconserved domains and their rela-
tionship to biological activity and antigenicity. Sot. Neurosci.
Abstr. 12: 214.
Greene, L. A. (1977a) Quantitative in vitro studies on the nerve
growth factor (NGF) requirement of neurons. I. Sympathetic neurons.
Dev. Biol. 58: 96-105.
Greene, L. A. (1977b) Quantitative in vitro studies on the nerve
growth factor (NGF) requirement of neurons. II. Sensory neurons.
Dev. Biol. 58: 106-l 13.
Greene, L. A., and G. Rein (1977) Release, storage and uptake of
catecholamines by a clonal cell line of nerve growth factor (NGF)
responsive pheochromocytoma cells. Brain Res. 129: 247-263.
Greene, L. A., and E. M. Shooter (1980) The nerve growth factor:
Biochemistry, synthesis and mechanism of action. Annu. Rev. Neu-
rosci. 3: 353-402.
Greiner, C. A. M., and G. Guroff (1986) Most catecholamine-con-
taining neural crest cells lack NGF receptors. Sot. Neurosci.
Abstr. 12: 1094.
Greiner, C. A. M., A. T. Lloyd, and G. Guroff (1985)
Developmental expression of nerve growth factor receptors in
primary cultures of neural crest cells. Sot. Neurosci. Abstr. I I:
937.
Greiner, C. A., A. T. Lloyd, and G. Guroff (1986) Ontogeny of
the nerve growth factor receptor in primary cultures of neural
crest cells. Brain Res. 391: 145-150.
Grillo, M. A. (1966) Electron microscopy ofsympathetic tissues.
Phar- macol. Rev. 18: 387-399.
Grillo, M. A., L. Jacobs, and J. H. Comroe, Jr. (1974) A
combined fluorescence histochemical and electron microscopic method
for studying special monoamine-containing cells (SIF cells). J.
Comp. Neurol. 153: 1-14.
Hervonen, H. (1974) Formaldehyde-induced fluorescence in the
sym- pathetic ganglia of the chick embryo in maturing organotypic
culture. Med. Biol. 52: 154-163.
Hervonen, H. (1975) Differentiation of sympathicoblasts in
cultures of chick ganglia. Anat. Embryol. 146: 225-243.
Hervonen, H., and 0. Eranko (1975) Fluorescence histochemical
and electron microscopical observations on sympathetic ganglia of
the chick embryo cultured with and without hydrocortisone. Cell
Tissue Res. 156: 145-166.
Hervonen, H., and L. Rechardt (1972) Formaldehyde-induced fluo-
rescence (FIF) in the sympathetic ganglia of chick embryos in
tissue culture. Stand. J. Clin. Lab. Invest. (Suppl. 122) 29:
69.
Hervonen, H., and L. Rechardt (1974) Light and electron
microscopic demonstration of cholinesterases of the cultured
sympathetic ganglia of the chick embryo. Histochemistry 39:
129-142.
Howard, M. J., and M. Bronner-Fraser (1986) Neural tube-derived
factors influence differentiation of neural crest cells in vitro:
Effects on activity of neurotransmitter biosynthetic enzymes. Dev.
Biol. 117: 45-54.
Iversen, L. L. (1967) The Uptake and Storage of Noradrenaline in
Sympathetic Nerves, Cambridge U. P., Cambridge.
Jacobowitz, D. M., and L. A. Greene (1974) Histofluorescence
study of chromaffin cells in dissociated cell cultures of chick
embryo sym- pathetic ganglia. J. Neurobiol. 5: 65-83.
Jacobs, M., Q. L. Choo, and C. Thomas (1982) Vimentin and 70K
neurofilament protein co-exist in embryonic neurons from spinal
gan- glia. J. Neurochem. 38: 969-977.
Johnson, G. D., R. S. Davison, K. C. McNamee, G. Russell, D.
Good- win, and E. J. Holborow (1982) Fading of immunofluorescence
during microscopy: A study of the phenomenon and its remedy. J.
Immunol. Methods 55: 23 l-242.
Jonakait, G. M., K. Markey, M. Goldstein, and I. B. Black (1984)
Transient expression of selected catecholaminergic traits in
cranial sensory and dorsal root ganglia in the embryonic rat. Dev.
Biol. 101: 5 l-60.
Kahn, C. R., and M. Sieber-Blum (1983) Cultured quail neural
crest cells attain competence for terminal differentiation into
melanocytes before competence for terminal differentiation into
adrenergic neu- rons. Dev. Biol. 95: 232-238.
Kahn, C. R., J. T. Coyle, and A. M. Cohen (1980) Head and trunk
neural crest in vitro: Autonomic neuron differentiation. Dev. Biol.
77: 340-348.
Kessler, J. A. (1985) Differential regulation of peptide and
catechol- amine characteristics in cultured sympathetic neurons.
Neuroscience 15: 827-839.
Kirby, M. L., and S. A. Gilmore (1976) A correlative
histofluorescence and light microscopic study of the formation of
the sympathetic trunks in chick embryos. Anat. Rec. 186:
437-450.
Landis, S. C., and P. H. Patterson (1981) Neural crest cell
lineages. Trends Neurosci. 4: 172-175.
LeDouarin, N. M. (1982) The Neural Crest, Cambridge U. P., Cam-
bridge.
Lee, V. M.-Y., and C. Page (1984) The dynamics of nerve growth
factor-induced neurofilament and vimentin filament expression and
organization in PC1 2 cells. J. Neurosci. 4: 1701-17 14.
Lever, J. D., and M. Presley (197 1) Studies on the sympathetic
neurone in vitro. Prog. Brain Res. 34: 499-512.
Levi-Montalcini, R,. and L. Aloe (1980) Tropic, trophic, and
trans- forming effects of nerve growth factor. Adv. Biochem.
Psychophar- macol. 25: 3-l 5.
Levi-Montalcini, R., and P. U. Angeletti (1968) Nerve growth
factor. Physiol. Rev. 48: 534-569.
Luckenbill-Edds, L., and C. van Horn (1980) Development of chick
paravertebral sympathetic ganglia I. Fine structure and correlative
histofluorescence of catecholaminergic cells. J. Comp. Neurol. 191:
65-76.
Mackey, H. M., R. F. Payette, and M. D. Gershon (1986)
Expression of serotonin, tyrosine hydroxylase and GABA in cultures
of neuro- genie cells from neural crest and branchial arches:
Effect of addition of the bowel. Sot. Neurosci. Abstr. 12:
1111.
Mains, R. E., and P. H. Patterson (1973) Primary cultures of
disso- ciated sympathetic neurons I. Establishment of long term
growth in culture and studies of differentiated properties. J. Cell
Biol. 59: 329- 345.
Mascorro, J. A., and R. D. Yates (1980) Paraneurons and
paraganglia: Histological and ultrastructural comparisons between
intraganglion paraneurons and extradrenal paraganglion cells. Adv.
Biochem. Psv- chopharmacol. 25: 201-214. - - -
Matthews, M. R., and G. Raisman (1969) The ultrastructure and
somatic efferent synapses of small granule-containing cells in the
su- perior cervical ganglion. J. Anat. 105: 255-282.
Maxwell, G. D., and D. S. Christie (1986) Properties of
catcholamine- positive cells which differentiate in neural crest
cultures. Sot. Neu- rosci. Abstr. 12: 1124.
-
The Journal of Neuroscience, November 1987, 7(11) 3763
Maxwell, G. D., and P. D. Sietz (1983) Expression of the
capacity to release [ZH]norepinephrine by neural crest cultures. J.
Neurosci. 3: 1860-1867.
Maxwell, G. D., and P. D. Sietz (1985) Development of cells con-
taining catecholamines and somatostatin-like immunoreactivity: Re-
lationship of DNA synthesis to phenotypic expression. Dev. Biol.
108: 203-209.
Maxwell, G. D., P. D. Sietz, and C. E. Rafford (1982) Synthesis
and accumulation of putative neurotransmitters by cultured neural
crest cells. J. Neurosci. 2: 879-888.
Maxwell, G. D., P. D. Sietz, and S. Jean (1984a)
Somatostatin-like immunoreactivity is expressed in neural crest
cultures. Dev. Biol. 101: 357-366.
Maxwell, G. D., P. D. Sietz, and P. H. Chenard (1984b)
Development of somatostatin-like immunoreactivity in embryonic
sympathetic ganglia. J. Neurosci. 4: 576-584.
Noden, D. M. (1978) Interactions directing the migration and
cyto- differentiation of avian neural crest cells. In Specificity
of Embryo- logical Interactions, D. Garrod, ed., pp. S-49, Chapman
and Hall, London.
NOIT, S. (1973) In vitro analysis of sympathetic neuron
differentiation from chick neural crest cells. Dev. Biol. 34:
16-38.
Oliver, D. L. (1985) Quantitative analyses of axonal endings in
the central nucleus of the inferior colliculus and distribution of
‘H-la- beling after injections in the dorsal horn of the cochlear
nucleus. J. Comp. Neurol. 237: 343-359.
Partlow, L. M., and M. G. Larrabee (1971) Effects of
nerve-growth factor, embryo age and metabolic inhibitors on growth
of fibers and on synthesis of ribonucleic acid and protein in
embryonic sympathetic ganglia. J. Neurochem. 18: 2101-2118.
Patterson, P. H., L. F. Reichardt, and L. Y. Chun (1975)
Biochemical studies on the development of primary sympathetic
neurons in cell culture. Cold Spring Harbor Symp. Quant. Biol. 40:
389-398.
Payette, R. F., G. S. Bennett, and M. D. Gershon (1984)
Neurofilament expression in vagal neural crest-derived precursors
of enteric neurons. Dev. Biol. 105: 273-287.
Price, J., and A. W. Mudge (1983) A subpopulation of rat dorsal
root ganglion neurons is catecholaminergic. Nature 304:
241-243.
Rothman, T. P., M. D. Gershon, and H. Holtzer (1978) The rela-
tionship ofcell division to the acquisition ofadrenergic
characteristics by developing sympathetic ganglion cell precursors.
Dev. Biol. 65: 322-34 I.
Shaw, T. J., and P. C. Letoumeau (1986) Chromaffin cell
heterogeneity of process formation and neuropeptide content under
control and nerve growth factor altered conditions in cultures of
chick embryonic adrenal gland. J. Neurosci. Res. 16: 337-355.
Shelton. D. L.. L. F. Reichardt. and E. M. Shooter (1986)
Isolation and sequencing of a clone of chicken genomic DNA encoding
a protein highly homologous to mouse beta nerve growth factor. Sot.
Neurosci. Abstr. 12: 1095.
Sieber-Blum, M. (1984) Fibronectin-regulated methionine enkepha-
lin-like and somatostatin-like immunoreactivity in quail neural
crest cell cultures. Neuropeptides 4: 457-466.
Sieber-Blum, M., and A. M. Cohen (1980) Clonal analysis of quail
neural crest cells: They are pluripotent and differentiate in vitro
in the absence of non-neural crest cells. Dev. Biol. 80:
96-106.
Sieber-Blum, M., S. R. Patel, and D. A. Riley (1986) In vitro
differ- entiation of quail neural crest cells into sensory-like
neuroblasts. Sot. Neurosci. Abstr. 12: 768.
Smith, J., and M. Fauquet (1984) Glucocorticoids stimulate
adrenergic
differentiation in cultures of migrating and premigratory neural
crest. J. Neurosci. 4: 2160-2172. - - -
Taoscott. S. J.. G. S. Bennett. and H. Holtzer (1981) Neuronal
pre- _ ~I ~ I cursor cells in the chick neural tube express
neurofilament proteins. Nature 292: 836-838.
Taxi, J. (1979) The chromaffin and chromaffin-like cells in the
auto- nomic nervous system. Int. Rev. Cytol. 57: 283-343.
Taxi, J., M. Derer, and A. Domici (1983) Morphology and
histophysi- ology of SIF cells in the autonomic ganglia. In
Autonomic Ganglia, L. Elfvin, ed., pp. 67-95, Wiley, Chichester,
UK.
Teitelman, G., T. H. Joh, and D. J. Reis (1978) Transient
expression of a noradrenergic phenotype in cells of the rat
embryonic gut. Brain Res. 158: 229-234.
Teitelman, G., H. Baker, T. H. Joh, and D. J. Reis (1979)
Appearance of catecholamine-synthesizing enzymes during development
of rat sympathetic nervous system: Possible role of tissue
environment. Proc. Natl. Acad. Sci. USA 76: 509-5 13.
Thoenen, H., and Y. A. Barde (1980) Physiology of nerve growth
factor. Physiol. Rev. 60: 1284-1335.
Thoenen, H., and J. P. Tranzer (1968) Chemical sympathectomy by
selective destruction of adrenergic nerve endings with 6-hydroxy-
dopamine. Naunyn-Schmiedebergs Arch. Pathol. 261: 27 l-288.
Tranzer, J. P., and H. Thoenen (1968) An electron microscopic
study of selective, acute degeneration of sympathetic nerve
terminals after administration of 6-hydroxydopamine. Experientia
24: 155-l 56.
Unsicker, K. (1973) Fine structure and innervation of the avian
ad- renal gland I. Fine structure and innervation of the adrenal
chromaffin cells and ganglion cell. Z. Zellforsch. Mikrosk. Anat.
145: 389-416.
Unsicker, K., B. Krisch, U. Otten, and H. Thoenen (1978) Nerve
growth factor-induced fiber outgrowth from isolated rat adrenal
chro- maffin cells: Impairment by glucocorticoids. Proc. Natl.
Acad. Sci. USA 75: 3498-3502.
Weber, K., G. Shaw, M. Osbom, E. Debus, and N. Geisler (1983)
Neurofilaments, a subclass of intermediate filaments: Structure and
expression. Cold Spring Harbor Symp. Quant. Biol. 48: 7 17-729.
Wechsler, W., and L. Schmekel (1967) Elektronenmikroskopische
Un- tersuchung der Entwicklung der vegetativen (Grenzstrang-) und
spi- nalen Ganglien bei Gallus domesticus. Acta Neuroveg. 30:
427-444.
Weibel, E. R. (1979) Stereological Methods, Academic, London.
Weston, J. A. (1970) The migration and differentiation of neural
crest
cells. Adv. Morphog. 8: 41-114. Wilson, S. P., K.-J. Chang, and
0. H. Viveros (1980) Synthesis of
enkephalins by adrenal medullary chromaffin cells: Reserpine in-
creased incorporation of radiolabeled amino acids. Proc. Natl.
Acad. Sci. USA 77: 4364-4368.
Wilson, S. P., M. M. Abou-Donia, K.-J. Chang, and 0. H. Viveros
(198 1) Reserpine increases opiate-like peptide content and
tyrosine hydroxylase activity in adrenal medullary chromaffin cells
in culture. Neuroscience 6: 71-79.
Winick, M., and R. E. Greenberg (1965) Chemical control of
sensory ganglia during a critical period of development. Nature
205: 180- 181.
Zacchei, A. M. (196 1) Lo sviluppo embrionale della quaglia
giap- ponese (Couturnix couturnix japonica T. e S.). Arch. Ital.
Anat. Em- briol. 66: 36-62.
Ziller, C., E. Dupin, P. Brazeau, D. Paulin, and N. M. LeDouarin
(1983) Early segregation of a neuronal precursor cell line in the
neural crest as revealed bv culture in a chemicallv defined medium.
Cell 32: 627- 638.